We examine the effect of academic entrepreneurship on collaboration between scientists by focusing on academic scientists who founded startups. We conceptualize collaborations of academic scientists as the formation, repeating, and exploitation of ties and propose that academic entrepreneurship has a negative effect on the formation of new ties and a positive effect on the repeating and exploitation of ties. We argue that academic entrepreneurship facilitates secretive behavior among scientists, increases entrepreneurial scientists’ concern with coordination cost, and shifts their attention from academic exploration to the success of the startup. Using a difference-in-difference analysis and coarsened exact matching on a longitudinal dataset of academic scientists in the life sciences and electrical engineering–computer science at five major U.S. research universities, we find robust support. Our findings highlight that academic entrepreneurship may unevenly enhance collaboration among academic scientists, thereby suggesting that academic entrepreneurship has an impact not only on entrepreneurial scientists themselves but also on the overall science community.